genericGANs

generic implementation of general adversarial networks in python

Getting Started

create a virtual environment and install package dependencies using the following code:

virtualenv .env --python python3.6
source .env/bin/activate
pip install -r requirements.txt

Running the MNIST example

if you would like to explore the MNIST linear GAN example as laid out in https://medium.com/ai-society/gans-from-scratch-1-a-deep-introduction-with-code-in-pytorch-and-tensorflow-cb03cdcdba0f

the program is run via

python lgan.py

Running the CIFAR10 example

included is a working example of deep convolutional GANs on the CIFAR10 dataset as laid out in https://github.com/diegoalejogm/gans

the program is run via

python dcgan.py

Visualizing the progress of the algorithm

this software uses the tensorboard implementation for pytorch, tensorboardX, to visualize the different parameters involved in training. while the program is executing, you may visualize the progress / development via tensorboard by running the following code:

tensorboard --logdir runs

and navigating to localhost:6006 in your web browser.

Happy hacking!

this project is under development, and will become extensible to data types of various shapes + allow for model flexibility